1
Atkins, P.J. (2010) Liquid Materialities: a History of Milk, Science and the Law
Farnham: Ashgate ISBN: 9780754679219 (hbk) and 9780754698197 (e-book),
334pp, http://www.ashgate.com/isbn/9780754679219
Chapter 1
A Material World
1.1 Material Quality
In both popular and scientific writing (in history, economics, sociology), a timeless and
objective definition of quality is commonly used, which appears to have been mobilised by an
outside observer. But in real markets, the quality of a product may diverge through time and
in different places.1
As Alessandro Stanziani points out in the quotation above, material quality in practical
reality is rather different from our representations of it. We idealise, we simplify, we
create object identities which help us to deal with a complex world, but which may also
serve a purpose, such as reassurance or the assignment of status or value. Stanziani’s
work is important because he refuses to take quality for granted and insists on a
genealogical perspective and a critical questioning of the very nature of a commodity. An
example is his observation that new products often begin as variants or imitations of
existing products, which in the eyes of some may be equated as plagiarism or
adulteration. He argues that choice and changes in taste are bound up with the
differentiation of quality, as is the diffusion of consumer behaviour among social classes
and income groups.
It is in this spirit of questioning that the present chapter will be devoted to
musings about material and material histories. It will introduce a number of perspectives
that are relatively recent in their development. It does not claim to be comprehensive
because a great deal in this general area has already been published, to the extent that
1 Stanziani 2007b: 209.
2
Graham Harman claims that ‘philosophy today is either materialist, or is intimidated by
materialism.’2
1.2 Metaphysics: Do You Take Milk with That?
Knowledge always refers to an object that is defined step by step through its history.3
The mission of this book is not philosophical, although its flavour is derived from a
number of ingredients that have philosophical connexions. It concentrates on emergent
material qualities, which, as noted in the Preface, distinguishes it from the efforts of most
food historians. In this section I will look at a history of relevant ideas at the
metaphysical level. Subsequent sections will then explore selected recent approaches to
the material.
Most sociologists of scientific knowledge have the very proper, mainstream desire
to unravel the epistemological problems that place limits on our knowing the world.
Because they study knowledge-generation in context, this leads to historically- and
spatially-contingent epistemologies. A good recent example is John Pickstone’s history
of science, technology and medicine, entitled Ways of Knowing. His account is a
summary of the intertwined histories of the gathering and making sense of knowledge in
natural history, experimental science, and the technoscience of industrial complexes.
Nikolas Rose espouses similar objectives in his ‘epistemology of assemblage.’ Rose’s
concern
is to reconstruct the epistemological field that allows certain things to be considered true at
particular historical moments, and the kinds of entities, concepts, explanations,
presuppositions, assumptions and types of evidence and argument that are required if
statements are to count as true … This approach does not seek to deny as such the
‘objectivity’ of knowledge, but to describe the ways in which objectivity is produced, and the
consequences of the production of objectivity.4
2 Harman 2009.
3 Gaston Bachelard cited in Bolduc and Chazal 2005: 81.
4 Rose 1999: xiiixv.
3
Along the same lines, part of the stimulation for the present project came from the
historical epistemology of Lorraine Daston and the historical ontologies of both Ian
Hacking and Ursula Klein. Taking historical epistemology first, this is the study of the
organization of knowledge through concepts such as objectivity, wonder, and error.5
Daston’s own definition is ‘the history of the categories that structure our thought, pattern
our arguments and proofs, and certify our standards of explanation.’6 I am mainly
interested in the material properties of milk, but its history, like the objects that appear in
Daston, is the result of competing forms of facticity: laboratory experiments, legal
ontologies, and legislative/regulatory/administrative ideas about what is ‘natural’ about
milk. These will be discussed in Chapters 3 to 9.
A good exemplar of historical epistemology is Mary Poovey’s wonderful book, A
History of the Modern Fact, where she traces the genealogy of the epistemological unit.
She argues that historical facticity, starting with the Baconian revolution of the
seventeenth century, concerns the failure of universals to ‘coalesce out of the common
experience of particulars.’7 Although there are traces here of Foucault’s genealogy,
Poovey develops her own vision of a contingent historical understanding of the
particulate nature of knowledge that is prior to the arrangement and deployment of facts
in discursive contexts, and she rejects the Foucaultian focus on the identification of
practices or events that constitute temporal ruptures. She uses double-entry book-keeping
as an example of the creation of modern facts, as the disinterested ‘nuggets of
knowledge’ that lubricated trade and allowed the construction of a rule-governed,
systematic mercantile knowledge base. Peter Miller summed this up in his observation
that it is ‘calculative practices that make the economy visible.’8
Daston’s own version of historical epistemology is about emergence. In her
Aristotelian metaphysical accounts, material objects are only seen as such by her
historical actors when certain conditions are met, most of all a solidification in the light
of their perceptual and knowledge-making resources. In these terms, artists have changed
5 Daston and Galison 2007, Schickore 2002.
6 Daston 1994: 282.
7 Poovey 1998: 8.
8 Miller 2001: 379.
4
their revelation of the human body from the flat, stylistic paintings of Cimabue in the
thirteenth century to, say, a Bill Viola video installation today. Likewise, food scientists
in the early nineteenth century experienced a sense of emergence when entirely new
aspects of the structure of foods came to light under the gaze of the microscope. In this
way, new worlds and new possibilities materialize, where they have not existed before,
although such objectivities are never fixed for more than a few years in the constant
churning of new ‘facts’ being born.
Daston and Galison explore this phenomenon of epistemic contextualization most
fully in their 500 page exegesis of scientific atlases. These books were made, they say, in
the contexts of distinct codes of epistemic virtue.9 What began as truth-to-nature in the
seventeenth century, became objectivity in the 1860s, and then turned into expertise in
the twentieth century. Each of these was part of the same spectrum, refracted differently
according to the epistemological circumstances of each era and each laboratory.
Observation, then, is relational and using the same nomenclature at different periods of
history does not conjure the same object. The bread of two hundred years ago, artisan-
made or domestic, was very different from the mass-produced, factory article of the same
name today. But then the consumers are also different, as is the flour, the yeast, and the
even the butter that is spread on it. A history of knowledge-making is needed to make
sense of this, and material histories have no meaning otherwise.
Daston and Galison show that objectivity was never quite achieved. It was always
just beyond reach, due to the technological constraints of laboratory equipment, of the
chemicals used in experiments, or of the accuracy possible with various forms of
mensuration. The recording of results and their interpretation were further uncertainties,
as were attempts to find a neutral means of conveying them to other scientists.
In an autobiographical moment, Daston recently recalled that Ian Hacking has
always disliked her historical meta-epistemology, as he calls it, preferring instead the
label ‘historical ontology.’10
What he has in mind for this is the study of ‘general and
organizing concepts and the institutions and practices in which they are materialized.’11
9 Daston and Galison 2007: 18.
10 Daston 2007: 806.
11 Hacking 2002: 12.
5
Taking up Hacking’s challenge, Ursula Klein and Wolfgang Lefèvre state that their
purpose for historical ontology as nothing less than a ‘new history of material objects in
general’ and, in particular, a reconstruction of eighteenth century scientists’ ontology of
materials through a study of their practices of identification and classification.12
Klein
and Lefèvre’s ‘thing history’, then, is the contextual history of material taxonomies.
Stuart Elden provides further clarification of the concept of historical ontology
through a reading of Foucault. He traces the roots of Foucault’s thinking to Nietzsche,
where a version of the genealogical method is to be found. But this does not amount to a
type of history. What Foucault is doing, rather like the later Heidegger, is searching for
‘different possible ways of being’ at a much deeper level, in other words ‘how what is
is’?13
The histories and geographies that arise from this are the result of a rejection by
him of the need to seek metaphysical universals.
Joseph Pitt finds a different form of words in a discussion of writers he calls ‘the
New Technists.’14
They share, he says, an interest in the knowledge embedded in
technological artefacts. This materialized version of historical ontology is close to my
own intentions, although my material ontogenesis makes no claims about metaphysics at
the deepest level. Rather, my search for substance histories involves: (a) the coming into
being of milk as a scientific object and commercial commodity; (b) its recognition as a
site of power; (c) its disciplining and harnessing as a lever of bio-politics; and (d) what it
can tell us about changing practices of knowledge. At the risk of sounding pretentious,
this amounts to ‘a history of the modern object.’
1.3 Material has Qualities; Quality is ‘Material’
The linguistic turn of post-structural social theory in the 1980s and 1990s meant a
privileging of the semiotic over the material, with texts, images and discourses coming to
dominate thinking.15
As a result, we were entertained by Roland Barthes’ accounts of
12 Klein and Lefèvre 2007: 1.
13 Elden 2001: 60, quoting Heidegger; Elden 2005: 356.
14 These include Bachelard, Galison and Hacking. See Pitt 2003.
15 Shove et al. 2007: 6.
6
mythologies such as ‘steak et frites’, awed by Michel Foucault’s ability to find evidence
of power everywhere, mesmerized by Deleuze’s cinematic vision of metaphysics, dazzled
by Baudrillard’s reading of signs, and, like as not, puzzled by Derrida’s claim that ‘there
is nothing outside of the text.’16
This pseudo monopoly of French theory was
complemented in the 1980s and 1990s by a worldwide surge of social constructionism,
which has also had a profound effect upon interpretations of human impacts on nature
(including food), as we will see later.
From the 1990s onwards the theoretical basis of critical opinion has widened,
with some scholars returning to philosophical basics in the work of twentieth century
giants such as Heidegger, who argued that the being of things is never fully present
before us. An object is therefore more than its appearance and more than its usefulness.17
There have also been moves towards a rematerialization of the social sciences, on the
premise that matter does matter despite it presently being under-represented in our
research.18
Scholars have been particularly keen to stress materiality as things-in-lives but
there has also been an encouraging post-human prioritization of the material thingness of
objects in their own right.
This new theoretical high ground was first occupied by ethnographers, notably
again in France but increasingly also in anglophone publications. Anthropologist Daniel
Miller has played a pivotal part in this emerging debate about materiality in a sub-field
known as material culture studies, with contributions, among many others, from
geographers Chris Philo and Peter Jackson, sociologist Tim Dant, and, in English
literature, Bill Brown.19
There is also a large and multivocal sociological and
ethnographic literature on things in Science and Technology Studies (STS)/Sociology of
Scientific Knowledge (SSK), which has been hugely influential in ontological reflections
upon objects.
The most thing-orientated of the social sciences, archaeology, has added its own
contribution to material studies. In a recent bout of soul-searching, attempts have been
16 Barthes 1972, Derrida 1976: 158.
17 Harman 2007: 1.
18 For a critical discussion of the rematerialization trend in geography, see Anderson and Wylie 2009.
19 Miller 1987, 1998, 2005a, Philo 2000, Jackson 2000, Dant 1999, 2005, 2008, Brown 2001, 2003, 2004.
7
made by archaeologists to shed both the vulgar materialism of their past, ‘in which
human behaviour was more or less shaped by non-human constraints,’ and also some of
the more extreme aspects of their ‘post-processual’ stress upon the experiential and the
phenomenological.20
Although Meskell indicates that archaeologists have some way to
go before they can be said to be pursuing the study of ‘object worlds’ in the full sense of
the material turn, they have at least made a start in what they call ‘symmetrical
archaeology.’21
It has not all been plain sailing for these various reborn materialists. A commonly-
heard complaint is that much of the new literature reduces its analysis of material to the
social. Matthew Kearnes, for instance, is particularly critical of the tendency to imply
‘that matter only matters in so far as it objectifies “social relationships” … rather than as
fundamentally independent and active.’22
Bjørnar Olsen has similar doubts:
To the extent that things are allowed to speak it is largely to bear witness to those human
intentions and action from which they themselves are believed to originate. Things may be
social, even actors, but are rarely assigned more challenging roles than to provide society with
a substantial medium where it can inscribe, embody and mirror itself.23
There are suspicions, then, that the material turn may involve some rebranding of
existing ideas rather than the radical break with the past that many were hoping for. An
example of the friction generated as a result of this disappointment is to be found in the
pages of the journal, Archaeological Dialogue, where a lively debate between Ingold,
Tilley, Knappett, Miller and Nilsson was published in 2007.
From our point of view, one solution to these uncertainties is to focus on material
qualities, prompting a big question: what is food quality? Although this might seem to be
a relatively straightforward, basic kind of question, soluble perhaps by asking producers
and consumers for their own criteria, the literature suggests a greater degree of
complexity and difficulty than is apparent at first sight. The idea of giving weight to
consumer views, for instance, seems to be relatively new, perhaps arising from the sheer
range of choice that is available in modern supermarkets, coupled with the fading of
20 Oestigaard 2004: 79.
21 Meskell 2004, Witmore 2007, Webmore 2007, Olsen 2003, 2007, Shanks 2007.
22 Kearnes 2003: 149.
23 Olsen 2007: 580.
8
concerns about organic deterioration that used to dominate thinking before the spread of
refrigeration technologies.
A number of recent papers identify the key dimensions of food’s material quality.
These are said to be: the naturalness of ingredients; qualities conferred by the method or
place of production; food safety and traceability; nutritional value; organoleptic qualities
and functionality; food’s biological quality; and quality in terms of certification.24
Although intended as commentaries on today’s food scene, these dimensions also have
some relevance to the historical period that is the subject of the present book, 1800 to the
present. I will endeavour to show this as the argument progresses. In addition, I will set
out to identify definitions of quality as they change through time, subject to a number of
influences. This is a key part of the book’s argument because insufficient attention has
been paid to the historical roots of our modern debate about food quality. Also, like
Harvey et al., I will argue that food quality is part of the more general question of how
people choose what to consume.25
Marie-Christine Renard puts this point succinctly:
‘Quality may … be defined as a product’s capacity to satisfy explicit or potential
consumer needs.’26
The notion of material quality has been undervalued in economics, to the extent
that Michel Callon calls it an ‘under-conceptualised and fragile notion.’27
In an attempt to
start a new debate, Musselin, Paradeise and their contributors have published an extended
deliberation on quality, in which they explore the disadvantages that its present neglect
brings for the study of markets and their mechanisms.28
Supply and demand are rarely
exactly balanced in real world markets, which are full of imperfections that result from
the asymmetric distribution of information. Attempts to solve such mismatches often
involve negotiations, not just about price, the usual consideration, but also about
24 Ilbery and Kneafsey 1998, Morris and Young 2000. See also Atkins and Bowler 2001.
25 Harvey et al. 2004.
26 Renard 2005, 421.
27 Callon 2005a, S94. See also Lévy 2002, Allaire 2004.
28 Musselin and Paradeise, 2005.
9
quality.29
It is therefore essential for us to know more about understandings of quality and
their origins.30
Callon et al. propose an ‘economy of qualities’ in which objects have shape, as a
moment in a never-ending process of production, and also a life, a career. This is
potentially applicable to food as it makes its way through a supply system, until it is
transformed into an edible dish and then starts a new journey as it is metabolically
absorbed or excreted as waste. Callon et al. then elaborate the definition of a good as a
bundle of qualities that establish its singularity. There is no essence, as such, just hybrid
characteristics, which, in a different combination, would make up a different good or a
variant of the same good.
Callon et al. see the qualities of goods as emergent rather than designed, or
revealed through ‘trials which involve interactions between agents (teams) and the goods
to be qualified.’31
Thus the taste and texture of a new dairy product is tested using
codified procedures and approved instruments, and, in the case of a well-established
product, there will also be routines to check that quality guidelines are maintained.32
Quality, in other words, cannot be taken for granted; it is forever threatening instability.
The characteristics of a good are not properties which already exist and on which information
simply has to be produced so that everyone can be aware of them. Their definition or, in other
words, their objectification, implies specific metrological work and heavy investments in
measuring equipment. The consequence is that agreement on the characteristics is sometimes,
in fact often, difficult to achieve. Not only may the list of characteristics be controversial
(which characteristics ought to be taken into consideration?) but so also, above all, is the
value to be given to each of them.33
In Callon’s economy of qualities the measurement of properties is important.34
Andrew Barry has elaborated the theme by talking of a government of qualities in which
various manifestations of the state invest in metrological regimes, such as the monitoring
29 Karpik 1989.
30 For more on the astonishing neglect of quality until recently, see Parrott et al., 2002.
31 Callon et al. 2002, 198.
32 For a Norwegian example, see Straete 2008.
33 Callon et al. 2002, 1989.
34 Callon et al. 2002. For a debate about Callon’s recent work, see Miller 2002, Fine 2003, and Callon 2005b.
10
and analysis of food quality.35
But these metrological regimes are frail. Any system of
monitoring that depends on point sampling has a high degree of uncertainty built in from
the outset, and then there are numerous means of cheating the process of sampling
available to the retailer. The results are therefore always difficult to interpret and may
have only a passing resemblance to the quality of the product reaching the consumer. In
addition, tests for compositional quality say nothing about the other dangers, such as the
spread of diseases like bovine tuberculosis.
Sensibly, Callon recognises the temporal dimension of product characteristics.
These change, with the material playing a role in the process of becoming, as does the
customer. These metamorphoses require further investment for what he calls
qualification-requalification, in order to stabilize and standardize goods for the market. A
new model of car, for instance, will undergo many design alterations before reaching the
market and then adjustments and redesigns before its career ends and it is replaced by a
new model. The customer will buy a variant suitable to her circumstances but, Callon
argues, part of the package will be intangibles such as the reputation of the manufacturer.
A parallel with the food economy is where a shopper smells and feels a mango before
purchase, but also takes on trust those material attributes, maybe its cultivar variety or
organic certification, which she cannot judge for herself on the spot.
Take milk, for instance. In most mammalian species it is white but close
inspection reveals a remarkable variation in its composition (Table 1.1). Even among
specialist dairy breeds of cattle, the term ‘milk’ in effect is a homonym for liquids
containing more or less butterfat and more or less solids-not-fat. So, we are entitled to
ask: ‘what is natural milk?’
One answer is that milk is such an astonishingly complex liquid that even now, at
the beginning of the twenty first century, it has still not given up the full story of its
organic chemistry.36
In the simplest of terms, it contains water, lipids, protein,
carbohydrate in the form of milk sugar (lactose), and also gases and minerals. It is an
emulsion of fat globules, a fine dispersion of casein micelles, a colloidal solution of
35 Barry and Slater 2002.
36 For a review of progress in dairy chemistry in the first half of the twentieth century, see Jenness 1956.
11
globular proteins and a colloidal dispersion of lipoprotein particles.37
The lipids comprise
neutral glycerides, free fatty acids, phospholipids, cerebrosides, gangliosides, sterols and
carotenoids.38
The most important protein is casein but also to be found are lactalbumin,
lactoglobulin, and fibrin, along with the enzymes (which act as organic catalysts)
peroxidase, reductases, lipase, phosphatase, catalyse, galactase and amylase.39
The
minerals (sometimes called ash in analyses of milk) are various salts of potassium,
sodium, calcium, magnesium, iron, sulphur, phosphorus and chlorine. There is also a
spectrum of vitamins and trace elements.
Table 1.1 The Average Percentage Composition of Different Milks
Selected species Dairy cattle breeds
Fat Protein Sugar Minerals Fat Protein Sugar Minerals
Bison 3.5 4.5 5.1 0.8 Ayrshire 4.14 3.58 4.69 0.68
Cow 3.7 3.4 4.8 0.7 Brown
Swiss
4.01 3.61 5.04 0.73
Donkey 1.4 2.0 7.4 0.5 Dexter 4.09 3.45 − −
Elephant 11.6 4.9 4.7 0.7 Guernsey 5.19 4.02 4.91 0.74
Goat 4.5 2.9 4.1 0.8 Holstein 3.55 3.42 4.86 0.68
Grey
seal
53.1 11.2 0.7 − Jersey 5.18 3.86 4.94 0.70
Horse 1.9 2.5 6.2 0.5 Red Poll 4.24 3.70 4.77 0.72
Human 3.8 1.0 7.0 0.2 Shorthorn 3.63 3.32 4.89 0.73
Polar
Bear
33.1 10.9 0.3 1.4 Simmental 4.03 3.38 − −
Rabbit 18.3 11.9 2.1 1.8 South
Devon
4.19 3.58 − −
Reindeer 16.9 11.5 2.8 − Sussex 4.87 9.31
Sheep 7.4 4.5 4.8 1.0 Welsh
Black
3.96 3.49 − −
Sources: Davis 1947, Fox and McSweeney 1998: 67, Jenness 1974: 91, Jenness 1999: 24,
National Milk Records 1980/1.
Further dimensions of milk composition have been revealed by dairy scientists.
For instance, in the early twentieth century it was understood that feeding regimes
influenced the quantity and quality of milk, but it was not dreamt that fodder type could
37 Walstra et al. 1999: 6.
38 Walstra et al. 1999: 50, Fox 1995.
39 Fox 1992.
12
be adjusted according to a desired breakdown into saturated, polyunsaturated and
monounsaturated fatty acids, as dictated by human health concerns.40
At the beginning of
the twenty-first century, microbiologists now think that milk contains roughly 100,000
types of organic molecules, most of which have yet to be identified and studied.41
It
remains much more of a mystery than we are perhaps willing to admit.
Although rules and standards encourage threshold notions of quality, where foods
either qualify or not for a special title or label, many of us nevertheless think of quality as
relative. All milk may be better in quality than was the case 100 years ago, perhaps across
every test one can think of in terms of cleanliness and freedom from disease or
contamination, but we still have ‘ordinary’ milk and ‘quality milk’ that can be purchased
in many forms. Once such premium milks become the market leaders, then, in their turn,
they will eventually be consigned to the category of ordinary or ‘regular.’ Meanwhile, of
course, we are all suspicious whether the ‘new, improved’ product is really any better
than before.
1.4 Performed Materiality
One of the most important trends in materiality studies has been a scrutiny of the
dynamic realm of practice in everyday life.42
As a result, the hardware of material
cultures is being reassessed in terms of its physical use. A particularly valuable
contribution has come from Elizabeth Shove and her collaborators.43
One of their projects
has been about interactive design in DIY and digital photography − the co-evolutionary
relationship between consumers and their objects that facilitates ‘distributed competence’
in the use of hammers and cameras, or indeed kitchen utensils, plastic cups and laptop
computers. This type of work overcomes a blind spot in research hitherto that assumes
the material of artefacts to be given. Following Heidegger’s tool analysis, Shove argues
that a hammer has no essential characteristics that can be said to define it and its kind;
40 Grummer 1991.
41 Singh and Bennett 2002: 1.
42 There is no space here to explore the important literature, mainly feminist, on bodily materialization. See, for
instance, the discussion of ‘bodies that eat’ in Probyn 2000.
43 Van Vliet et al, 2005, Shove et al. 2007, Hand and Shove 2007.
13
and when it is picked up and used for hammering, it is the practice that tells us more than
the implement.
In a similar vein, Emma Roe has worked on the processual emergence of the
material qualities of foods, in other words ‘the process of some thing becoming food’ and
‘the doing of eating.’44
She does so by observing the embodied practices of consumers
who have a material attunement to eating organic food and her interest is in the
‘meaning-making event when a thing, a foodstuff, becomes food, becomes eaten.’45
Her
focus groups discussed understandings of the food that they had bought and prepared, and
individually they then made video diaries in the kitchen. Unlike most consumption
studies, Roe’s methodology enabled an account to be written that ‘does take the “guts” of
eating practices seriously in terms of the diversity of individual eating practice, the
consequential multiple meanings of edibility that arise, and the potential to make ethical
and political material connections.’46
The work of Kevin Hetherington also has appeal. His book Capitalism’s Eye
deals with the experience of commodities in department stores, exhibitions and in other
forums of spectacle, and his article on touch and textures is also interesting for its radical
take on materiality.47
For the latter he interviewed a visually impaired lady whose touch
knowledge of objects in museums he takes to be performative rather than
representational. Her claim is that ‘when I am touching something there is no “me” and
the object I am touching. It is just the object. So the me disappears ... for me it’s just
touching, identifying with the actual thing there.’48
Hetherington concludes that one way
in which place is generated is from this kind of unmediated experience, especially
through routine, and this is confirmed, as he puts it, ‘by the praesentia found in the dirt of
time.’49
Although Hetherington does not discuss food as such, we could see taste in the
same terms as a form of proximal knowledge. After all, the mouth is a major site of the
44 Roe 2006a: 465 and 470, emphasis as in original.
45 Roe 2006b: 105.
46 Roe 2006a: 479.
47 Hetherington 2003, 2007.
48 Hetherington 2003: 1934.
49 Hetherington 2003: 1943.
14
performance of our intimate connexion with the world, hungrily incorporating it and
sensually experiencing it. Each mouthful is an experimental test of organoleptic quality
and a way of monitoring whether our expectations have been met. Taste, then, can be
seen as an embodied quality of the material. It is both a physiological process with a
distinct set of purposes and culturally specific.50
Thus, similar foods may be valued
differently by various cultural groups and the ability to distinguish subtly different
flavours, for example in wine, can function as a class marker.51
Matters are complicated
further when taste information is used to evaluate the degree of excellence of food. Doing
so is predicated on a hierarchical evaluation of food characteristics that reflects power.52
Such hierarchies are culturally defined, just as the characteristics that underpin them are
culturally mediated.
One of my arguments in this book will be that the material of foodstuffs may be
understood in ways that are emergent in practice. Tasting, consuming and metabolising
are elements of this but maybe the strongest observation for our purpose is that food is a
site for ontological politics. This terminology is derived from the ethnography of objects
and their associated knowledges undertaken by Annemarie Mol, who asserts that
‘ontology is not given in the order of things … instead, ontologies are brought into being,
sustained, or allowed to wither away in common, day-to-day, sociomaterial practices.’53
In her world of hospitals, there are diseases and instruments and knowledges associated
with their diagnosis and treatment, and these knowledges are enacted through practical,
material events. She advocates a radical empirical philosophy in which the focus is on
what she calls ‘praxiographies’ of the multiplicity of material reality. She pointedly
criticizes the constructionism of STS/SSK, arguing that it ignores the fragility of object
identity because ‘matter isn’t as solid and durable as it sometimes appears.’54
Her
principal contribution is in helping us to understand the slipperiness of making diagnoses
of medical conditions such as atherosclerosis but her radical, empirical style is equally as
50 Stassart and Whatmore, 2003.
51 Guthman 2002: 295 and 300.
52 Gronow 1997.
53 Mol 2002: 6.
54 Law and Mol 1995: 291.
15
applicable to food studies as to disease. 55
The present book is pitched at a different scale
and is historical in intent, so a reproduction of Mol’s method is impossible, but it seems
to me that the underlying aim of her ontological politics is relevant. In what follows, I
will attempt to demonstrate the elusiveness of knowing milk, which in the period under
study was only partially overcome in the practice of science, commerce and the law.
1.5 Relational Materiality
There has been neglect in the theory of food studies of the material of production, a
surprising omission. Becky Mansfield is prominent among those attempting to redress
this imbalance, in her case by stressing biophysicality in the food geographies of fish
(catfish and surimi).56
The task she sets herself is to understand how definitions of quality
emerge in specific economic contexts. Rather than making a general distinction between
nature and society or arguing for socially constructed quality, she looks at individual
production networks in order ‘to understand how specific aspects of what we call “the
natural world” participate in particular interactions.’57
Her idea of quality is different
because she argues that it arises from assemblages of practices within commodity
chains.58
These chains have particular histories that have established sets of relations and
the commodities are differentiated as a result.
Mansfield uses the spatial metaphors of distancing and entanglement to
understand the trajectories of fish products in their commodity chains. She looks in
particular at surimi, a Japanese fish paste which is a mixture of fish protein and starch.
Because of its physical characteristics, surimi has a wide range of functionality, including
products such as ‘krab’ sticks (artificial crab). It can be made from a number of different
fish species and so can be produced in many countries. Quality is based on fish biology,
processing technologies, mouthfeel (texture, chewyness), whiteness, and the nature of the
55 Mol 1999, Mol and Mesman, 1996, Harbers et al, 2002.
56 Mansfield 2003a, 2003b, 2003c, 2003d.
57 Mansfield 2003a: 10.
58 For more on assemblages as a Deleuzian concept, see De Landa 2006.
16
final product. Such issues can be dealt with at the local level and overseen by
transnational food corporations, making this a food with a global reach.
Mansfield’s conclusion is interesting, that ‘economic activities materialize
meanings.’59
She insists that quality is relational. It is ‘neither a subjective judgment
(what different people like), nor an objective measure (the characteristics of a
commodity), but instead it is produced within relations of commodity production and
consumption.’60
Jane Bennett’s interest in the major electrical outage in North America’s
major power grid in August 2003 is similar in many ways to Mansfield’s foregrounding
of assemblages, known by Deleuze as agencements. These are arrangements of actants
with a dynamic interconnexion. Bennett is willing to concede agency − ‘thing-power’ −
not only to individuals, such as artefacts and technologies, but also to collectives of
humans and nonhumans that ‘pulse with energies, only some of which are actualized at
any given time and place.’61
Support for the idea of that material quality is an emergent feature of
interconnectedness may be found in Bill Cronon’s book Nature’s Metropolis, which
presents a fascinating picture of early America.62
Part urban history, part environmental
history, and part a history of the frontier, it also makes a contribution to food history
because Chicago was a focus for both the grain and livestock trades of the prairies and
west. Cronon’s breakthrough is in showing that there were strong links between the
success and growth of Chicago and the exploitation of its wide hinterland. This took the
form of the makeover of the landscape to such an extent that one could say that a field in
western Illinois was just as much a part of Chicago as Madison Street or Michigan
Avenue. Maintaining the fiction of a ‘natural’ landscape is not easy under such
circumstances and so it is much easier to see the city and the countryside as fused in a
relationship that was mutually constitutive. This was reinforced by the transport and
telegraph networks and by a role that eventually made Chicago one of the world’s great
centres of commodity trade. The stock yards and the Board of Trade became international
59 Mansfield 2003c: 330.
60 Mansfield 2003a: 1112.
61 Bennett 2005: 461, Bennett 2004.
62 Mansfield 2003a: 13, Cronon 1991.
17
symbols of raw material capitalism. Two keys to success were the adoption of a system
of grading wheat and the linked growth of a futures market. The development and
stabilising of quality was, then, a vital component in the success of Chicago’s main
agricultural industries.
London performed a similar gathering function in nineteenth century Britain. Its
docks, railway termini and vast food wholesale markets made it a nodal hub for the
Empire. But, before we get carried away with whiggish enthusiasm for growth and the
wonders of complex system structure, it is important to point out that we are not
assuming without challenge that the formation of long distance linkages is necessarily
‘good.’ One might be forgiven for taking this message from some of the published
material, which evaluates, say, logistical efficiency or the formation of social capital. On
the contrary, as we will see, there are networks of shared values that are at odds with
those of mainstream society. In its early stages, the system of bringing railway milk to the
big cities in the 1850s and 1860s, for instance, almost certainly facilitated large-scale
cheating because no-one had a responsibility for standards and there was a reduction of
traceability in a network that had a high index of anonymity. Extending this logic, we
might go on to argue that the present-day ‘crisis’ of food identified by Beck in his Risk
Society is really no more than a readjustment of consumers’ perceptions away from the
quality they attributed to their local butcher or supermarket and towards a networked idea
of quality, where responsibility is lodged at the more diffuse level of the nation state or a
supranational block such as the European Union. In other words, there has been a scale
jump in the spatial texture of provision, which ‘the system’ has not yet been able to
accommodate.
Peter Jackson finds the various accounts of commodity chains and systems to be
‘too linear, too mechanistic and too focused on the simple metric of length as opposed to
other issues such as complexity, transparency or regulation.’63
By comparison,
‘commodity circuits’, exemplified in the work of geographers such as Ian Cook and Phil
Crang, at least have the virtue of having ‘no beginning and no end’ and they acknowledge
that ‘origins are always constructed.’64
Commodity circuits ‘attend to cultural inflected
63 Jackson et al. 2006: 132.
64 Cook et al. 1998.
18
relationships between production, circulation and consumption.’65
They also pay
attention to the knowledges and understandings of commodities held by farmers, traders
and consumers, for instance in what is called by Kopytoff the ‘biography of things.’66
There is a link here with the materiality literature because of calls to take the circulating
food commodities themselves seriously. Ian Cook, for instance, urges us to ‘get with the
fetish’ and to ‘follow the thing.’67
For the purposes of the present book, both chains and circuits are useful
conceptual vehicles in as much as they consider the contractual arrangements and
institutional governance of food, which helps to provide a basis for understanding the
origins and driving mechanisms of food quality. Some authors suggest mediation
between the different approaches: Leslie and Reimer, for instance, between systems of
provision and commodity circuits. An important degree of flexibility in any chain study is
also an openness to product-specificity. This is vital for work on food, where the
characteristics of, say, wheat and horticultural products to a certain extent determine how
they are handled.
Recent work in geography, economic sociology and institutional economics
suggests that the fundamentals of food quality may be traced to rules and conventions, as
set out, for instance, in the theory of conventions, a major thrust of action-oriented
French pragmatism.68
Starting in the 1980s, conventions theorists have been interested in
the social relations associated with production and exchange.69
As they remind us, there
is much uncertainty between actors in food chains, but this can be coped with through
conventions – formal or informal agreements that arise out of situations that vary
according to the product in hand. Conventions emerge over time through the successful
repetition of particular actions and may have their roots in tacit agreements, without
written rules. They are shaped by the material environment of action and in turn become
65 Braun 2006a: 645.
66 Kopytoff 1986.
67 Cook et al. 2004, Cook 2004.
68 For introductions in English, see Wagner 1994, 2001, Wilkinson 1997, Favereau and Lazega 2002. A
comprehensive retrospective French language collection was published in two volumes by Eymard-Duvernay in 2006.
69 Storper and Salais 1997.
19
a guide for future action and a form of legitimation that can be discussed and modified.70
Institutions may also arise in the long run, for instance to deal with disputes, and
conventions may overlap with each other or be replaced, in time, by contractual
obligations between parties. Even these formal arrangements are, nevertheless, social
constructs, and, as a result, they are heterogeneous in their composition and deployment.
Early on in the conventions literature Storper and Salais looked at manufacturing
industry. Their ambitious project was to reformulate the structure-agency debate in
economic geography in terms of the pragmatics of action. They pursued the ‘archaeology
or genealogies of economic situations’, showing a sensitivity to the emergence of ‘worlds
of production.’ For them, the product is a central feature of their approach, because it
‘embodies and thus realizes the potentialities of the resources of action’, but their stress is
not so much product input and manufacturing costs as upon the ordered practices that
arise from long-run routines, agreements, assumptions and expectations.71
Eymard-Duvernay recognizes three types of coordination linking quality to the
availability of information. First, where quality is unknown, unreliable, or unstable, the
trust that is established over long periods between buyer and seller is a substitute. Second,
questions of quality can be referred to a neutral adjudicator, such as a public analyst in
the case of doubts about the chemical composition of or physical characteristics of foods.
Third, concerns about quality may be judged in terms of the health of the consumer, as
was frequently the case with the infectious diseases spread by milk.72
Conventions theory has been extensively used to understand the development of
quality in food networks. According to Jonathan Murdoch, it has advantages when
comparing networks, whereas actor network theory is more introspective because of its
inductive approach.73
Conventions are also of interest because they encapsulate the
interpersonal world of contacts in the world of speciality and high quality foods.
Conventions theory by now has a richness that encompasses several interpretations of the
role of conventions.74
The most appropriate for our purposes are conventions that were
70 Thévenot 2006: 112, Lévy 2002: 263.
71 Storper and Salais 1997: 1415.
72 Eymard-Duvernay 1989.
73 Murdoch 1998, Murdoch and Miele 1999.
74 Woolsey Biggart and Beamish 2003.
20
customary and, as such, were treated informally as social rules. It was not until the late
nineteenth and early twentieth centuries that normative expectations among traders were
replaced by legally binding regulations and, even then, there remained an element of
negotiation with nature in the minds of most actors.
In their ‘worlds of production’ analysis, Storper and Salais speak of the
‘personality’ of products, by which they mean a commodity’s profile of specialization,
substitutability, perishability, quality differentiation, its potential for technological
development, and so on. In Britain’s dairy industry it is possible to identify three worlds
of production in the late nineteenth century, reduced to two by the end of the twentieth.
These were: (a) urban and suburban milk producers, called cowkeepers, who employed
intensive systems, often with cattle confined permanently, to their dark and poorly
ventilated shed; (b) country producers who had the good fortune to be located close to the
ever-expanding railway network, and who were able to find a city market for their output
up to 200 miles away; (c) specialist dairy farmers who were primarily producers of butter
and cheese, but who from time to time were called upon to supply ‘accommodation milk’
in times of general shortage.
The main feature of the conventions associated with the milk trade from the early
nineteenth century until the 1960s was that the perishability of the product, and its
vulnerability to infection with disease, meant that it had to be produced and transported in
a timely and cleanly manner if it was to travel efficiently along networks to the
customer’s doorstep. In the very simplest of terms, we can identify some of the principles
behind drinking-milk in our period that eventually came to be commonly accepted after
decades of negotiation:75
‘Milk as it came from the cow’;
Daily delivery, fresh not sour;
Stability in compositional quality over period of contract;
Freedom from disease and dirt;
Wholesaler responsible for supply chain management, i.e. balancing differences between
supply and demand;
‘Invalid’ and ‘infant’ milk from special cows.
75 See Sylvander and Biencourt 2006.
21
Milk seems to fit the ‘industrial’ convention of Boltanski and Thevenot. It was
produced intensively in several specialised areas, with the just-in-time flexibility of
additional accommodation milk when the supply/demand balance was disrupted for one
reason or another. Table 1.2 shows the conventional responses made by different actors
to the problems that arose on a regular basis. These amount to a lived critique of the
system, which could only be maintained as it was by the cynicism and fatalism of the
consumer. By the early twentieth century there were economies of scale in production,
processing and transport, making the industry attractive to investment.
Table 1.2 Typical Reactions to Milk Problems by Actors in the Period c. 18501940
Producer Trader Analyst Medical Officer Consumer
Insufficient or
unreliable
supply
Increase
herd
‘Accommodation
milk’ or
adulteration
Test N/A Use condensed
milk
Not rich
enough
Stet Colorants Test Sale of Food
and Drugs Acts
Use whole
rather than
skim milk
Sour Cool milk Preservatives Test N/A Complain to
retailer
Diseased Sell
diseased
cattle
Pasteurization Test Examine cows Buy heat
treated milk
Dirty NIRD
methods
Filter Test Examine
cowshed
Buy certified
milk
1.6 Hybrid Materialities
The 1990s saw a gradual shift away from the chain metaphor of food systems and from
political economy, both of which had served agri-food scholars well over the previous
twenty years.76
There had already been the work of Michel de Certeau on the texture of
daily life and of James Scott on ‘the weapons of the weak’, which had suggested that the
juggernaut of international capitalism and its associated cultural traits could not roll
without resistance into every corner of daily life.77
Granovetter’s timely intervention had
also stressed the power of the social in shaping market structures and processes. In
76 Busch and Juska 1997.
77 De Certeau 1984, De Certeau et al. 1998 , Scott 1985.
22
addition, the post-structural turn in social theory was sending shock waves through the
cosy certainties of the social sciences. When actor network theory came on to the scene in
the 1980s, the time was right for its rapid acceptance and spread. Starting in science
studies as an ethnographic means of understanding the working of scientists and their
laboratories, it was soon applied in human geography and in sociology. Some agri-food
writers have put it to work, if not to paint a coherent picture, then at least for its radical
unsettling of many previously unchallenged assumptions.
Bruno Latour, the ‘amodern’ magician of objects takes a bow at this point. He has
become an obligatory passage point for STS/SSK engagement with the material because
actor network theory, as propounded by Latour, Callon and others, is the non-
constructionist materialism that has attracted most attention overall. It asserts that objects
have hybrid qualities and so avoids the classic dualism of nature and society, the Achilles
heel of much constructionist writing. Actor network theory has attracted a great deal of
attention: praise and criticism in almost equal measure. One the one hand, it has been
called by one enthusiast ‘the most promising philosophy of our time.’78
On the other
hand, one of Latour’s most important recent books, Politics of Nature, was described by a
broadly sympathetic reviewer as ‘300 pages of abstruse verbiage.’79
The foundational concept of actor network theory is that socio-natural hybridity is
expressed in quasi-objects. In this spirit, Bakker and Bridge talk of ‘the mongrel nature of
the world’, where ‘we navigate a world made up of radically incommensurable things,
suturing them together as we go.’ Milk, for instance, can be seen as a blend of human and
non-human agency ready to take its place in Latour’s ‘parliament of things.’ This is a
forum in which material is no longer mute, no longer reliant upon scientists, or anyone
else, to speak on its behalf. Foodstuffs, which are simultaneously naturalized and
socialized, deserve their place in the textbooks of food history on their own terms. Their
stories will be biographies, but also, because they contribute control, partly
autobiographies.
This idea of hybrids − objects in motion blended with human action – promotes a
relational and distributed view of materiality in which ‘the competencies and capacities
78 Harman 2009.
79 Castree 2006: 164.
23
of “things” are not intrinsic but derive from association.’80
The principal expression of
this relational and dispersed agency is the network. Networks may include humans;
milking machines; ‘inscription devices’ such as milk yield records and laboratory
analysis results; cows; and materials like railway wagons and delivery carts. These are
joined together in alliances that vary in strength and persistence. Some coordination and
stability may be provided by ‘centres of calculation’ such as the official Government
Laboratory in London, but this cannot be taken for granted and networks face constant
challenge and change. For example, the long-distance milk supplies that came to London
in winter in the late nineteenth century, which were only possible if every connexion and
technology worked smoothly, in summer had to be reviewed because hot weather
reduced the souring time and threatened to unsettle the whole network. But actor network
theory is not really about spatially discrete networks in the sense of a sewage or a telephone
system, nor is it structural as with a social network. A better metaphor is said to be the
rhizomes of Deleuze and Guattari, which in my lighter moments I think of as couch grass
theory.
Hybrids and networks are in a constant state of becoming, as a result of the process
of translation, through which actants come to represent ‘a multitude of others by defining
and linking their identities in increasingly simplified and fixed forms.’81
This may be by
persuasion, seduction or force. One result of translation is the creation of ‘mixtures
between entirely new types of beings, hybrids of nature and culture.’82
Another result is the
creation of what we might call ‘associational truth’, which is the product of alliances in
networks. According to Latour ‘the strongest reason always yields to reasons of the
strongest.’83
This, by the way, is not a sell out to relativism or constructionism but a
demonstration that the germ theory, which dominated medical science for decades from the
1870s onwards, derived its strength from an accumulation of alliances, including many
weak ties, which were generated and exploited by Pasteur in his campaign to establish the
truth of his science. In retrospect we know that he was ultimately successful, but at the time
80 Bakker and Bridge 2006: 16.
81 Shiga 2006: 41.
82 Latour 1993: 10.
83 Latour 1988: 186.
24
this was by no means guaranteed. He faced rivals for the explanation of the spread of
anthrax and the souring of milk, and in each case he also faced resistance from the materials
that he worked with.
Critics argue that actor network theory escapes socio-historical contingencies, and
Latour responds that history is made of networks, for instance the much cited work by
Thomas Hughes on the historical growth of electricity supply systems, or Braudel on the
growth of the networks of capitalism.84
This is a very superficial presentation of actor network theory. While I am more
than willing to concede the vital folding of the non-human world into agency, and that
‘societies do not construct nature as they please’, I cannot go further.85
The strong
symmetry between human and non-human actors that is proposed by Latour, and
brilliantly taken up in her own way by Sarah Whatmore, seems to me in its extreme
versions to be counter-intuitive.86
I have more sympathy with the call of Pels for a ‘weak
asymmetry.’87
Whatmore’s hybrid geographies are balanced nature-culture-spaces,
whereas mine emphasize the hybrid milk-in-society. My rather different type of narrative
is structured around human knowing and intervening within the limits of material
potentials and constraints. While actor network theory works for me as a kind of agora for
ideas about objects, it leaves too little room for cultural politics.
1.7 Material Qualities are Processual
Matter comes to matter through the iterative intra-activity of the world in its becoming.88
We come now to the section which most closely reflects the approach to materiality
adopted in this book. I say that guardedly because, as the book unfolds, the reader will
see that this initial theoretical reflection provides a departure point but not a destination.
Further commentaries will be necessary from time to time to provide support for a text
84 Hughes 1983, Braudel 1982.
85 Castree 2001: 17.
86 Whatmore 2002.
87 Pels 1996: 296.
88 Barad 2003: 823.
25
that is largely about food. The special qualities of this vital part of our lifeworld have not
been considered in depth by material philosophers, and the conceptual depth of food
history, the discipline, is insufficient to carry us through the story that I want to tell.
Martin Heidegger’s exceptionally powerful philosophy can help us, especially
that to be found in his Being and Time. To begin with, he suggests that apprehending an
object and its material substance is impossible in the sense of immediate phenomenology
because the very being of that object is embedded in time and, of course, we cannot press
a pause button in order to allow us to walk around it and see it from every angle. The
properties of milk are always at least partially absent and only become ‘present at hand’
when someone encounters it with one of the commonly agreed needs for ‘milk.’ Objects
to Heidegger are not bundles of properties, but are formed in a system of relations, which
in turn is temporally contingent.89
Milk’s thingness, then, is not just a matter of its whiteness, or its nutritional value,
or its usefulness in making my tea drinkable. Yes, under certain circumstances it does
have these and other material and functional characteristics. But the full materiality of
milk is always partially withdrawn. Heidegger did not have in mind histories in the sense
of clock time but this should not discourage an interest in genealogy and ontogenesis. We
can now see that a study of milk’s facticity will be far more than a matter of observation,
even under the controlled conditions of a laboratory. It will be multidimensional. Even at
the end of this book, devoted as it is to only one material, we will be left thinking that
there is much more to say. Indeed there is, but even the subsequent three volumes of the
series I am intending will still not be enough to exhaust the possibilities.
Another of Heidegger’s key concepts, that of dwelling, has been explored by Tim
Ingold. Through this he devises an anti-constructionist account of objects and landscapes
that is based on a mutually co-constitutive, dialectical process in which people are a part
of the ecology and ‘take part in generating their own form and the form of their
surroundings.’90
Ingold argues that society’s features are a co-production with its
surroundings, which presents opportunities but also restrictions according to time and
place. In fact, Ingold’s confident interventions make him a good person to start this last
89 My interpretation of Heidegger relies heavily upon Harman 2007.
90 Berglund 1998: 70.
26
lap of Chapter 1. This is because he is uninhibited in his comments about the material
turn. He is critical of social constructionism and deeply sceptical of work on
‘materiality’, which, according to him, is ‘expounded in a language of grotesque
impenetrability on the relations between materiality and a host of other, similarly
unfathomable qualities, including agency, intentionality, functionality, sociality,
spatiality, semiosis, spirituality and embodiment.’ Instead, he is interested in the
processual properties of materials, which to him need to be contextualized in temporal
and spatial terms.91
This is because the
very property is a condensed story. To describe these properties means telling the stories of
what happens to them as they flow, mix and mutate … The properties of materials … are not
attributes but histories.
Ingold refuses to concede agency to materials but sees them as active ‘because of
ways in which they are caught up in these currents of the lifeworld.’ His materials are not
bystanders: ‘far from being the inanimate stuff typically envisioned by modern thought,
materials … are the active constituents of a world-in-formation.’
Some of Ingold’s readers may interpret his statements in a realist sense. This is
not his intention but one can see why his thoughts on processual properties might appeal,
for instance, to historical ontologists working in the philosophy of medicine. An example
is the assertion by Smith and Brogaard that ‘each substance is a bearer of change.’ This
provides a philosophical basis in logic for a dynamic vision of biomedical ontologies, and
they illustrate its potential in their article ‘Sixteen Days’, which seeks to identify the
point when a growing embryo can be called human.92
Predictably, this is controversial,
but Barry Smith’s work generally seeks to demonstrate the value of rigorous ontologies
in medicine and philosophy. Rooted in Spinoza’s Ethics, it is helping biologists to think
through the significance of spatial boundaries. Smith has worked extensively on social
and biomedical ontologies, and also on the formal ontology of Edmund Husserl, which,
he suggests, amounts to a general theory of objects.93
Smith’s epistemology has much in
91 The quotations are from Ingold 2007.
92 Smith and Brogaard 2003: 47.
93 Smith and Smith, 1995. It is important to note that Husserl was Heidegger’s senior colleague but the eventual
intellectual break between them was irreparable.
27
common with Geographical Information Science, which is widely used in geography,
history and other disciplines where dynamic cartography is important.
Despite the power of Smith’s biological ontogenesis, a genetic concept of change
is problematic for our purposes because it is pre-programmed with certain limits. The
story I wish to tell about the history of food quality has fewer constraints and might have
turned out very differently. This prompts questions about the nature of material and
technological change that have answers in two very different literatures, both depending
to an extent upon systems thinking.
The first is the so-called ‘path dependence’ theme of economic history, which has
been energised principally by the efforts of Paul David. Although backed by a statistical
vocabulary familiar to econometricians, David’s argument is about the nature of
historical narrative and his message is that ‘history matters’ in explaining present
economic circumstances. A path dependent system is one that evolves as a result of its
own history, and so events bear the traces, often distant memories, of the past. The
purpose is to counter the assumptions of classical economics that human behaviour is
subject to timeless laws, which have tended to lead to a low priority in explanation for
spatio-temporal contingency and for events that lead to what David calls a ‘forking in the
road.’ None of this will shock historians, even those reluctant to consider counterfactual
versions of their stories. So what is special about path dependence?
David’s formulation is interesting, not just because it posits alternative stable
states of an economic (or social, cultural, political) system. He also deals with
‘regrettable’ branchings, which, because of irreversibility, may mean the adoption of sub-
optimal technologies, institutions, and possibly lead to market failure. The example
given, repeated now so often that is has become a cliché, is that of the QWERTY
keyboard dating from the 1870s.94
I am typing on one right now (very slowly with one
finger) although I know that it is not ergonomically the best solution to minimize my
effort. Design inertia is not an adequate explanation of this curious survival. We must
turn instead to the concept of lock-in, through which a system state is stabilised, without
immediate prospect of moving forward and seeking alternatives.95
This may be because
94 But see Liebowitz and Margolis 1990.
95 Arthur 1989.
28
of institutional self-reinforcement and/or dynamic increasing returns through learning
mechanisms. Other well-known examples of this phenomenon include Edison’s direct
current versus Westinghouse’s alternating current, and the triumph of the VHS video
system over Betamax. It seems that decisions made early in a technological cycle may
have a lasting impact despite their sub-optimal nature. Chance may play a part in this but
system features are also crucial.
Building on this, Martin and Sunley claim that there has been what we may call a
‘precedent turn’ in economic geography, economic sociology and evolutionary
economics in recent years.96
This has amounted to taking history seriously and has arisen
from a general and profound disquiet with the assumptions of rationality implicit in the
general equilibrium model of economic behaviour. Path dependence might help to
explain the origin and persistence of regional clusters of firms, and maybe also various
forms of negative regional lock-in that manifest themselves as aspects of technological
and institutional rigidity that are difficult to break when a new competitor region appears.
Martin and Sunley point to existing ideas that have some resemblance to this literature,
for instance those on palimpsest economic landscapes in the work of Doreen Massey and
to a certain extent David Harvey. But they are also critical of the many unresolved
aspects of path dependence. It is not clear, for instance, how new paths are established
and old ones maintained, and what their relationship is with economic evolution.
The second literature of technological change is based on the work of the French
philosopher, Gilbert Simondon. Simondon was interested in inventions and technical
developments, for instance engines used in the motor car industry. Design changes and
practical adjustments mean the gradual emergence of improved technologies and it is
sometimes difficult to identify when a new object has replaced the old one. What is the
character of the object and how does it become stabilised? These are the sorts of
questions asked by Simondon, for instance in his 1958 book Du Mode d’Existence des
Objets Techniques.97
As Isabelle Stengers has pointed out, Simondon was seeking
answers to the chicken and egg problem because his objects rarely have a point of origin,
96 Martin and Sunley 2006. See also Boschma and Martin 2009; Grabher 2009.
97 This has been translated by Ninian Mellamphy in a thesis for the University of Western Ontario (1980). See also
Dumouchel 1992.
29
nor even clear cut stages of development.98
To use a simile from Simondon himself, they
are rather like a coral reef that grows slowly, building upon its own previous
achievements. But, to be recognisable objects, they must have self-referential
independence, as in the ignition of a diesel engine or the whiteness of milk.99
Although very little of Simondon is in translation, we are fortunate to have
summaries of his arguments provided by Massumi, Mackenzie, Toscano and Stiegler, and
a number of French interpretations and special journal issues have also been published.100
Simondon’s key concept is that of ‘individuation’, as expounded in books such as
L’Individu et sa Genèse Physico-Biologique (1964) and L’Individuation Psychique et
Collective (1989). Simondon’s ontological vision is one of becoming, in which the
material and the forces that shape it are seen as a pooled presence.101
He sought a
philosophy capable of grasping developments, of genesis, not of identities or substances,
which meant returning to a pre-Socratic view of existence. He seems to have been
influenced by Bergson and Whitehead, and in turn Deleuze was influenced by
Simondon.102
Like Simondon, Whitehead saw objects as historical events. His often quoted
example is Cleopatra’s Needle, the Egyptian obelisk sited by the River Thames in
London, which he claimed is a happening made up of ‘event-particles.’103
To him, to
Simondon, and to Deleuze, ‘all “beings” are just relatively stable moments in a flow of
becoming-life.’104
Deleuze later developed Simondon’s individuation further, for instance
in the concept of ‘machinic assemblages’ that he formulated with Guattari.105
Simondon understood technical objects ‘as evolving composites of relations
rather than in terms of function, use, material or form.’106
This means that they acquire a
concretized character as the result of an historical process in which human intervention is
98 Stengers 2004.
99 Dumouchel 1992: 414.
100 Stiegler 1998, Massumi 2002, Mackenzie 2002, 2005, 2006, Toscano 2006, Combes 1999, Barthélémy 2005, 2006.
101 Chabot 2003: 77, Chabot 2005.
102 Stengers 2002, Halewood 2005.
103 Whitehead 1920.
104 Colebrook 2002: 125.
105 Deleuze and Guattari 1987.
106 Mackenzie 2006: 200.
30
both facilitated and constrained by the potentiality of the object. An indicator of this is
internal coherence.
Like Whitehead, Simondon sees an individual, not as an entity, but as a
process.107
Henning Schmidgen calls this ‘serial being’ and for Ajit Nayak it is the
‘indivisible continuity of reality.’108
Following this lead, for us ‘becoming milk’ is a way
of seeing the means by which humans and nature have formed a hybrid with a degree of
stability. Simondon’s ‘ultimate phenomenon’ is an assemblage of individuation and
individual in which the material object we observe is only one of its phases, the most
recent.109
The successive stages of individuation are accomplished by a process of
transduction: the ontogenetic repetition or modification of form.110
Simondon’s ideas are different from those of the social constructionists. For him,
change emerges in the process of individuation, not as a result of the interaction between
subject and object. This is because individuation is preceded by what he calls a ‘pre-
individual’, which has the potential for change.111
An example is the DNA of the
fertilized egg that produced each of us. This provided a set of genetic potentials but our
actual form was a matter of unfolding. This continues now and will continue in the future
as we individuate further in the process of ageing. The individual is never final.
One aspect of our bodily individuation is what we eat. We take on the form of that
food through the process of nutrition. The scholastics saw this but, rather than embracing
it, they developed a theory of substantial change, in which food loses its shape in the
body due to the digestive process.112
But bioarchaeologists can infer much about past
diets because certain groups of food leave their signatures in the chemistry of bone
collagen. The skeleton, then, is a record of food consumption that can be recovered by
stable isotope and mineral analyses.113
107 Shaviro 2007.
108 Schmidgen 2005: 17, Nayak 2008: 177.
109 De Beistegui 2005: 118.
110 Mackenzie 2002. For more on transduction, see Dodge and Kitchin 2005.
111 Mackenzie 2006.
112 Chabot 2003: 87.
113 Unfortunately, the technology as it stands does not allow us to distinguish between the consumption of meat and
dairy produce. Müldner and Richards 2006.
31
1.8 Conclusion
There are, of course, many other avenues for exploring material and materiality. A
favourite of mine is Henri Focillon’s Vie des Formes, which was first published in 1934
and is still in print. Tom Conley, in his translator’s preface to Deleuze’s book, The Fold,
points to the relevance of Focillon’s work.114
Focillon’s essay was a theoretical
justification of his major work, The Art of the West, which rejected the tradition of closed
genealogies and grid-like classification of styles of painting or architecture. On the
contrary, rather than watertight periodizations, Focillon found artistic hyperlinks across
the centuries that produced eclectic hybrids and non sequiturs where they might not have
been expected. Focillon’s approach was influenced by Bergson, and, later, Focillon’s
student, George Kubler, continued the interest in the flow of history in his The Shape of
Time.115
While making parallels between art history and food history is stretching
explanation to its limits, particularly since Focillon and Kubler were both interested in
form rather than material, we can nevertheless admire their conclusion that ‘things are
materialized attempts to solve problems.’116
Also Kubler offers us a sharp and timely
reminder about the role of the scholar in material histories:
The ‘shape of time’ is not immediately given in the things themselves, but results from the
work of the historian. The series and sequences into which he or she groups the forms of
things and problems retrospectively alter the arrangements of things that were hitherto
accepted. As a consequence, the historian changes even the forms themselves. Against this
background, the history of experimentation might be read as a succession of shapes, the
production of which sets in motion a cascade of retroactive re-shapings.117
So, sources of inspiration for the present volume are potentially many and varied.
I suggest that so far we have formulated four conceptual seeds that can be grown on and
planted out. First, no matter what the labelling on their work – historical epistemology or
historical ontology – I take heart from those authors, all heavily influenced by Foucault,
114 Deleuze 2006: ix.
115 Molotiu 2000, Kubler 1962.
116 Max Planck Institute for the History of Science 2006: 6.
117 Max Planck Institute for the History of Science 2006: 6.
32
who have taken a genealogical stance on knowledge. Starting with Chapters 2 and 3, we
will learn how intellectual context was crucial for the way in which farmers, experimental
scientists, politicians, and the attentive public saw and understood milk. This tells us a
great deal about ways of thinking but also sheds light on the material itself. What was
milk after all?
Second, readers will not be surprised that social and material constructionism has
influenced my thinking to a certain extent. It has been a powerful movement in the last
twenty years and its logic remains deeply embedded in much post-structuralist writing.
But constructionism has passed its apogee and post-human theory and the various forms
of materialism have chipped away at its credibility. Rather than reproducing the social-
natural dichotomy yet again, scholars are trying to find ways to let the material speak for
itself. Also, some new writing is stressing, in retrospect, how foolish it was to ascribe
god-like, nature-changing powers to individuals, institutions, and whole societies who
patently lack the resources, expertise, technologies, and political willpower to intervene,
for instance, in the major socio-environmental problems of the day, such as climate
change. The inherent resistance and messiness of socio-natures is now a focus of study,
and in science studies Andrew Pickering’s mangle of practice looks much closer to a
practical framework of analysis than much of the earlier, frankly naïve constructionism.
Third, our account of non-representational theories in general has thrown up a
number of ideas to take forward. Work on practice, especially Mol’s discussion of the
fragility of object identity, will be of help, although obviously my historical
interpretations along these lines are in a different register to her praxiographies. The
hybridities of actor network theory will also provide some energy to my story, although I
must repeat that borrowing some of Latour’s concepts does not mean that we have to
swallow his metaphysics.
Fourth, and finally, there are several processual literatures that strike a chord.
Ingold’s reading of Heidegger provides a helpful insight into thingness, as does Smith’s
Husserlian ontogenesis of bodily form. But it is Simondon’s philosophy that, for me, is
the most exciting. It is potentially relevant to the present book if we count milk as form of
technology. As we will see, it was a fluid that was ‘made’ by producers. They bred cattle
that gave large volumes of low fat milk and they designed feeding regimes for maximum
33
output, a development so profound that Manuel De Landa sees it as a new form of
biological history.118
There were then issues about the legitimacy of manipulating the
constituents of milk as it made its way through the supply chain. Much of our discussion
will be about the means of detecting fraud and imposing sanctions. Finally, public policy
with regard to a legal definition of milk will be another of our themes. Although a
Simondian analysis is possible for items of food and drink such as bread, meat, wine, it
seems to me that milk is the most appropriate commodity.
In summary, then, this book is about a product which, though it retained the
undifferentiated and unqualified name of ‘milk’ throughout the period under study, was
highly variable. As a technological object, its material form went through bursts of rapid
change and long periods of stability. It has existed in a state of what Simondon called
‘metastability’, exhibiting tension, never static. To treat milk in the nineteenth and
twentieth centuries as a homogeneous, natural foodstuff would be inappropriate. Its
individuation, its change, is the subject of this book.
118 De Landa 2000: 1634.